A virtual assistant that can record the words spoken and show the corresponding webpage.
Live training by multiple speakers is supported.
New words can be added for new as well as old speakers.

Started with the command based version which can be found here

1. Open Yugi.sln file in Visual Studio 2010.
2. Press F5 or ctrl+F5 for building and running the solution.
3. You can start with the existing model.
4. If you want to go back to the new-born version, press "start/reset" button. It will delete all extra trainings and will keep only the default words which were added in the beginning of this project.
5. Then you can try live testing. Speak the word within 3 seconds. The corresponding webpage will be displayed.
6. For live training, first write the word you want to train. If you're adding a new word, you'll have to record 10 times continuously.

1. Home page:
   

2. Recording page:
   

3. Wiki result:
   

Fork this repository, make required changes and create a pull request.

A final project submitted for the course fulfilment requirement of Speech Processing at IITG. Developed by Rohan Jaiswal and Prateekshya Priyadarshini.

1. YES/NO Detection
   A Simple project to detect two words i.e. Yes and No based on the fricatives.
2. Vowel Recognition
   An application to detect the vowels spoken.
3. Finding Optimal Codebook
   A mini project to find the optimal codebook using K-Means and LBG ALgorithms.
4. Command Version of Yugi